A series of experiments were carried out in Purdue University’s Boeing/AFOSR Mach6 Quiet Tunnel, to understand the amplification of streamwise vortices across a separated region in a quiet flow regime. Streamwise vortices were induced on the upstream end of an axisymmetric model consisting of a 7-degree half-angle cone, a cylinder, and a 10-degree flare. The instabilities were seeded using a pre-existing set of roughness inserts, with small, discrete roughness elements. The elements varied in spacing, height, and number of elements. The model was aligned to near 0.0 degree angle of attack. <div><br></div><div>The streamwise, Gortler-like instabilities travelled across the separated region onto the flare, where they were measured with pressure transducers and infrared thermography. The amplification of the instabilities was measured at a variety of Reynolds numbers, under both quiet and conventional noise flow. The results were compared to those of a smooth insert. Heat transfer results showed a streaking pattern, with a peak in heating visible in the streak. Heat flux increased linearly with Reynolds number. If transition was induced, the heat flux would begin to decrease. Power spectral density measurements of the pressure fluctuations indicated that the region within the streak contained two notable instabilities, one between 70 and 150 kHz, and one between 200 and 250 kHz. Transition was only measured in the spectral content in the region on the flare where a ”filling in” of streaks was visible in heat transfer results. Heat flux increased in an nonlinear manner with increasing roughness height. </div><div><br></div><div>The streak positioning and peak heat flux showed a high sensitivity to small, uncontrollable changes in run conditions throughout. Heat transfer results were largely repeatable for small angles of attack, less than 0.1 degrees. The streaks shifted slightly in width and position for angles of attack near 0.1 degrees. Small changes in the streak positioning and heat transfer magnitude were seen in repeatability runs; this is mostly attributable to small changes in initial run conditions. </div>